Making things should be as cheap as imagining them.

In a handful of years, AI has driven the cost of producing software toward zero. Writing code is no longer the scarce thing — judgment, taste, and knowing what to build are. That shift is not a prediction. It already happened, in the open, fast enough that we're still adjusting to it.

The physical world has not moved the same way. Making a physical object still demands scarce, expensive expertise long before anything reaches a machine. And the first wall you hit isn't manufacturing — it's geometry. Turning an intention into a precise, manufacturable model you can actually change is slow, and it gates everything downstream. Nothing after it gets cheap until that step does.

That is exactly where AIDA begins. Describe a part — or hand it a drawing or a photo — and get back a living, parametric model, with every key dimension exposed as something you can drive. Not a thinner CAD tool; a way past the modelling step entirely. We chose the unglamorous, foundational problem on purpose, because it's the one that unlocks the rest.

Geometry is the first step, not the last. Simulation folds the laws of physics into the loop, so a design is refined against how it will actually behave — not just how it looks. From there, the path runs toward turning a model into something a machine can make. Each step compounds on the one before it, which is why the order matters.

Design is one half of what a physical thing costs. Physical labour is the other. General-purpose robotics is on a trajectory that rhymes with software's — and when the cost of doing physical work falls the way the cost of writing software already has, the second half gives way too.

When both designing and making things become cheap, the cost of a physical object falls toward the cost of imagining it. That is the direction of an economy of abundance — a post-scarcity direction, where fewer things are scarce because they're hard to make. I don't think that's utopian. I think it's the honest extrapolation of a curve we are already on.

A claim that large is only worth anything if you earn it one shipped step at a time. So we don't lead with the destination — we lead with the part that works today. We started at the bottleneck. The rest follows from there.